Tuning Wiki

This is a historical revision. Read the current page or return to its history.

Page organisation in this revision
Page type
Technical article
Subject
Emissions control
Parent
Section level
Applies to
No ECU-specific applicability
Tool groups
None
Manufacturers
None

Changes from revision 1 to revision 1

Removed text is struck through; added text is highlighted.

The control unit does not weigh the filter

Soot and ash are masses, in grams, that the control unit computes. The only direct measurement on the brick is the pressure drop, and pressure drop is not a mass until it is read against the exhaust flowing through the filter and against the temperature of that gas. A differential pressure that looks high at idle can be a modest soot load. The same number at high exhaust flow can be a nearly empty filter. Every load model is a way of turning that family of curves, plus what the engine has been doing, into a number the regeneration strategy can use.

Three arrangements are in production. A diagnostic tool often labels the result soot mass in all three.

Pressure only

The oldest practical model stores a map of expected pressure drop against exhaust flow, or against a stand-in for flow such as air mass and engine speed, for a clean filter and for several soot loads. The measured drop is placed on that map and reported as a load. Temperature corrections account for the viscosity and the density of the exhaust.

There is one number. Ash and soot both move it. After a regeneration the combustible soot is gone and the ash plug is not, so the reported load no longer falls to the new-filter line. The control unit then regenerates more and more often as the car ages, because it still sees soot, or a technician resets the value and the pressure sensor immediately walks it back up. A pressure-only strategy has no separate ash service limit. The filter is condemned when the empty pressure, or the distance, reaches the threshold chosen for that calibration.

A soot model, checked by the pressure sensor

Later control units integrate soot. At each moment the engine-out rate is looked up from speed, injected quantity, air mass and EGR, because those are what set how much carbon the spray makes. That rate is added to a soot-mass integrator. When the filter is hot enough for oxidation, a burn rate is subtracted. The burn rate depends on temperature, oxygen, and whether nitrogen dioxide is available, which is the coating and any fuel-borne catalyst.

The pressure sensor is the check, not the only source. Measured drop is compared with the drop the model predicts for the current soot mass at the current flow and temperature. Inside a band, the integrator is trimmed so it cannot drift for ever. Outside the band, the two disagree and a plausibility fault is stored. A hose blocked with water, a sensor offset, or a brick that has melted and lost its restriction all fail this comparison. So does a filter whose ash has lifted the baseline while the soot integrator still thinks the wall is clean.

A completed regeneration reduces the soot integrator to a small residual. It does not subtract ash, because the strategy has nowhere to put ash. Distance and time triggers sit beside the mass trigger. Many calibrations also request a regeneration after a set distance if the exhaust can be made hot, so a filter is not left holding a moderate load for months of gentle driving.

Soot and ash as separate masses

The fuller model keeps two integrators.

Soot mass is the combustible load. It rises with engine-out soot and falls with oxidation and with a commanded regeneration. It is the number that requests a burn. Ash mass is the incombustible load. It rises with the oil the engine is calculated to have consumed, with the metal the additive pump has dosed, and with mileage. It does not fall when a regeneration finishes.

Ash is also adapted from the pressure that remains after a regeneration the soot model considers complete. At a known exhaust flow, a new filter has a known drop. The same flow through a filter full of ash plugs reads higher. That offset is written into the ash mass, and the soot map is then read on top of it. The next soot cake is judged from the pressure above the ash line, not from the pressure above zero. This is the arrangement in which a regeneration can succeed, the soot mass can return to its residual, and the car can still be a service case because the ash mass or the residual pressure has reached its limit.

The two masses produce different actions. Soot requests heat. Ash requests a workshop visit: a clean or a new brick, then a reset of the stored ash so the model matches the part just fitted. Leaving the old ash mass in memory makes a new filter look full. The control unit then limits torque, or demands a regeneration that cannot reduce a number which is not soot.

When ash monitoring is switched off

The ash integrator is a calibration choice. Some control units were never given one, and they are the soot-model case above. Others have the integrator and a switch, and the switch is off in that software. Production updates have been issued that stop adapting ash where the counter was climbing faster than the filter was really filling and was condemning bricks that still had capacity. With the monitor off, the stored ash mass stays where it was, often at zero, and only soot is adapted.

The pressure rise from real ash then has to go somewhere. If the soot integrator is still trimmed by the pressure sensor, the ash slowly appears as soot that a regeneration cannot clear, and burns become more frequent with age. If the soot mass is reset hard after every regeneration and the pressure check is wide, the ash stays invisible until the back-pressure, the high-load smoke, or a temperature the regeneration can no longer control makes it obvious. A car whose ash monitor is off still makes ash. The software has stopped counting it.

Reading the diagnostic value needs that distinction. A soot mass of zero after a drive that never got hot is a model that has been reset, or a monitor that is not running. A soot mass that returns to the same high value within a minute of a completed burn is the ash line being reported as soot. The pressure, the flow and the temperature at that moment are what separate the two.

What the model cannot see

A cracked or melted wall loses restriction. The pressure falls, the model reports a low load, and the soot is leaving through the hole. An efficiency check, where the calibration has one, compares the pressure behaviour with what a filtering wall would do, and stores a fault when the brick looks too empty. A channel face plugged with a sheet of ash does the opposite: the pressure is high while the soot mass, if it is not being trimmed, still looks moderate. The sensor hoses, the exhaust temperature sensors and a plausible air-mass reading are part of the same calculation. An air-mass meter that under-reads moves the flow the model uses, and the load comes out wrong on a filter that is mechanically fine.

Rendered revisions

Open each version to inspect formatting, tables and photographs.

Revision 1: DPF load models

The control unit does not weigh the filter

Soot and ash are masses, in grams, that the control unit computes. The only direct measurement on the brick is the pressure drop, and pressure drop is not a mass until it is read against the exhaust flowing through the filter and against the temperature of that gas. A differential pressure that looks high at idle can be a modest soot load. The same number at high exhaust flow can be a nearly empty filter. Every load model is a way of turning that family of curves, plus what the engine has been doing, into a number the regeneration strategy can use.

Three arrangements are in production. A diagnostic tool often labels the result soot mass in all three.

Pressure only

The oldest practical model stores a map of expected pressure drop against exhaust flow, or against a stand-in for flow such as air mass and engine speed, for a clean filter and for several soot loads. The measured drop is placed on that map and reported as a load. Temperature corrections account for the viscosity and the density of the exhaust.

There is one number. Ash and soot both move it. After a regeneration the combustible soot is gone and the ash plug is not, so the reported load no longer falls to the new-filter line. The control unit then regenerates more and more often as the car ages, because it still sees soot, or a technician resets the value and the pressure sensor immediately walks it back up. A pressure-only strategy has no separate ash service limit. The filter is condemned when the empty pressure, or the distance, reaches the threshold chosen for that calibration.

A soot model, checked by the pressure sensor

Later control units integrate soot. At each moment the engine-out rate is looked up from speed, injected quantity, air mass and EGR, because those are what set how much carbon the spray makes. That rate is added to a soot-mass integrator. When the filter is hot enough for oxidation, a burn rate is subtracted. The burn rate depends on temperature, oxygen, and whether nitrogen dioxide is available, which is the coating and any fuel-borne catalyst.

The pressure sensor is the check, not the only source. Measured drop is compared with the drop the model predicts for the current soot mass at the current flow and temperature. Inside a band, the integrator is trimmed so it cannot drift for ever. Outside the band, the two disagree and a plausibility fault is stored. A hose blocked with water, a sensor offset, or a brick that has melted and lost its restriction all fail this comparison. So does a filter whose ash has lifted the baseline while the soot integrator still thinks the wall is clean.

A completed regeneration reduces the soot integrator to a small residual. It does not subtract ash, because the strategy has nowhere to put ash. Distance and time triggers sit beside the mass trigger. Many calibrations also request a regeneration after a set distance if the exhaust can be made hot, so a filter is not left holding a moderate load for months of gentle driving.

Soot and ash as separate masses

The fuller model keeps two integrators.

  • Soot mass is the combustible load. It rises with engine-out soot and falls with oxidation and with a commanded regeneration. It is the number that requests a burn.
  • Ash mass is the incombustible load. It rises with the oil the engine is calculated to have consumed, with the metal the additive pump has dosed, and with mileage. It does not fall when a regeneration finishes.

Ash is also adapted from the pressure that remains after a regeneration the soot model considers complete. At a known exhaust flow, a new filter has a known drop. The same flow through a filter full of ash plugs reads higher. That offset is written into the ash mass, and the soot map is then read on top of it. The next soot cake is judged from the pressure above the ash line, not from the pressure above zero. This is the arrangement in which a regeneration can succeed, the soot mass can return to its residual, and the car can still be a service case because the ash mass or the residual pressure has reached its limit.

The two masses produce different actions. Soot requests heat. Ash requests a workshop visit: a clean or a new brick, then a reset of the stored ash so the model matches the part just fitted. Leaving the old ash mass in memory makes a new filter look full. The control unit then limits torque, or demands a regeneration that cannot reduce a number which is not soot.

When ash monitoring is switched off

The ash integrator is a calibration choice. Some control units were never given one, and they are the soot-model case above. Others have the integrator and a switch, and the switch is off in that software. Production updates have been issued that stop adapting ash where the counter was climbing faster than the filter was really filling and was condemning bricks that still had capacity. With the monitor off, the stored ash mass stays where it was, often at zero, and only soot is adapted.

The pressure rise from real ash then has to go somewhere. If the soot integrator is still trimmed by the pressure sensor, the ash slowly appears as soot that a regeneration cannot clear, and burns become more frequent with age. If the soot mass is reset hard after every regeneration and the pressure check is wide, the ash stays invisible until the back-pressure, the high-load smoke, or a temperature the regeneration can no longer control makes it obvious. A car whose ash monitor is off still makes ash. The software has stopped counting it.

Reading the diagnostic value needs that distinction. A soot mass of zero after a drive that never got hot is a model that has been reset, or a monitor that is not running. A soot mass that returns to the same high value within a minute of a completed burn is the ash line being reported as soot. The pressure, the flow and the temperature at that moment are what separate the two.

What the model cannot see

A cracked or melted wall loses restriction. The pressure falls, the model reports a low load, and the soot is leaving through the hole. An efficiency check, where the calibration has one, compares the pressure behaviour with what a filtering wall would do, and stores a fault when the brick looks too empty. A channel face plugged with a sheet of ash does the opposite: the pressure is high while the soot mass, if it is not being trimmed, still looks moderate. The sensor hoses, the exhaust temperature sensors and a plausible air-mass reading are part of the same calculation. An air-mass meter that under-reads moves the flow the model uses, and the load comes out wrong on a filter that is mechanically fine.

Revision 1: DPF load models

The control unit does not weigh the filter

Soot and ash are masses, in grams, that the control unit computes. The only direct measurement on the brick is the pressure drop, and pressure drop is not a mass until it is read against the exhaust flowing through the filter and against the temperature of that gas. A differential pressure that looks high at idle can be a modest soot load. The same number at high exhaust flow can be a nearly empty filter. Every load model is a way of turning that family of curves, plus what the engine has been doing, into a number the regeneration strategy can use.

Three arrangements are in production. A diagnostic tool often labels the result soot mass in all three.

Pressure only

The oldest practical model stores a map of expected pressure drop against exhaust flow, or against a stand-in for flow such as air mass and engine speed, for a clean filter and for several soot loads. The measured drop is placed on that map and reported as a load. Temperature corrections account for the viscosity and the density of the exhaust.

There is one number. Ash and soot both move it. After a regeneration the combustible soot is gone and the ash plug is not, so the reported load no longer falls to the new-filter line. The control unit then regenerates more and more often as the car ages, because it still sees soot, or a technician resets the value and the pressure sensor immediately walks it back up. A pressure-only strategy has no separate ash service limit. The filter is condemned when the empty pressure, or the distance, reaches the threshold chosen for that calibration.

A soot model, checked by the pressure sensor

Later control units integrate soot. At each moment the engine-out rate is looked up from speed, injected quantity, air mass and EGR, because those are what set how much carbon the spray makes. That rate is added to a soot-mass integrator. When the filter is hot enough for oxidation, a burn rate is subtracted. The burn rate depends on temperature, oxygen, and whether nitrogen dioxide is available, which is the coating and any fuel-borne catalyst.

The pressure sensor is the check, not the only source. Measured drop is compared with the drop the model predicts for the current soot mass at the current flow and temperature. Inside a band, the integrator is trimmed so it cannot drift for ever. Outside the band, the two disagree and a plausibility fault is stored. A hose blocked with water, a sensor offset, or a brick that has melted and lost its restriction all fail this comparison. So does a filter whose ash has lifted the baseline while the soot integrator still thinks the wall is clean.

A completed regeneration reduces the soot integrator to a small residual. It does not subtract ash, because the strategy has nowhere to put ash. Distance and time triggers sit beside the mass trigger. Many calibrations also request a regeneration after a set distance if the exhaust can be made hot, so a filter is not left holding a moderate load for months of gentle driving.

Soot and ash as separate masses

The fuller model keeps two integrators.

  • Soot mass is the combustible load. It rises with engine-out soot and falls with oxidation and with a commanded regeneration. It is the number that requests a burn.
  • Ash mass is the incombustible load. It rises with the oil the engine is calculated to have consumed, with the metal the additive pump has dosed, and with mileage. It does not fall when a regeneration finishes.

Ash is also adapted from the pressure that remains after a regeneration the soot model considers complete. At a known exhaust flow, a new filter has a known drop. The same flow through a filter full of ash plugs reads higher. That offset is written into the ash mass, and the soot map is then read on top of it. The next soot cake is judged from the pressure above the ash line, not from the pressure above zero. This is the arrangement in which a regeneration can succeed, the soot mass can return to its residual, and the car can still be a service case because the ash mass or the residual pressure has reached its limit.

The two masses produce different actions. Soot requests heat. Ash requests a workshop visit: a clean or a new brick, then a reset of the stored ash so the model matches the part just fitted. Leaving the old ash mass in memory makes a new filter look full. The control unit then limits torque, or demands a regeneration that cannot reduce a number which is not soot.

When ash monitoring is switched off

The ash integrator is a calibration choice. Some control units were never given one, and they are the soot-model case above. Others have the integrator and a switch, and the switch is off in that software. Production updates have been issued that stop adapting ash where the counter was climbing faster than the filter was really filling and was condemning bricks that still had capacity. With the monitor off, the stored ash mass stays where it was, often at zero, and only soot is adapted.

The pressure rise from real ash then has to go somewhere. If the soot integrator is still trimmed by the pressure sensor, the ash slowly appears as soot that a regeneration cannot clear, and burns become more frequent with age. If the soot mass is reset hard after every regeneration and the pressure check is wide, the ash stays invisible until the back-pressure, the high-load smoke, or a temperature the regeneration can no longer control makes it obvious. A car whose ash monitor is off still makes ash. The software has stopped counting it.

Reading the diagnostic value needs that distinction. A soot mass of zero after a drive that never got hot is a model that has been reset, or a monitor that is not running. A soot mass that returns to the same high value within a minute of a completed burn is the ash line being reported as soot. The pressure, the flow and the temperature at that moment are what separate the two.

What the model cannot see

A cracked or melted wall loses restriction. The pressure falls, the model reports a low load, and the soot is leaving through the hole. An efficiency check, where the calibration has one, compares the pressure behaviour with what a filtering wall would do, and stores a fault when the brick looks too empty. A channel face plugged with a sheet of ash does the opposite: the pressure is high while the soot mass, if it is not being trimmed, still looks moderate. The sensor hoses, the exhaust temperature sensors and a plausible air-mass reading are part of the same calculation. An air-mass meter that under-reads moves the flow the model uses, and the load comes out wrong on a filter that is mechanically fine.